Jump to content

  • Log In with Google      Sign In   
  • Create Account


Member Since 16 Oct 2013
Online Last Active Today, 05:10 AM

Posts I've Made

In Topic: target selection strategies

Yesterday, 05:08 PM

If the behavior "makes sense" to the player, we have succeeded.

It's worth noting that what makes sense to the player is something we can (to a certain extent) teach the player during the game.


Imagine a few scenarios:

You're surrounded by enemies, and they all suddenly take cover without warning just before a thrown grenade explodes.

You stealthily take out a guard in a far-off location. As you're looting the corpse, 15 other enemies show up.


Both of those could feel very unfair to the player -- the AI is cheating!


Now imagine the same scenarios, but imagine they are telegraphed to the player.

One of the enemies shout "Fire in the hole!" right before the group of enemies take cover.

You hear radio chatter on the corpse's walkie-talkie, which sounds more and more distressed, before the rescue party shows up.


The AI didn't need to change, but now the game feels a lot more clever, because we've been more explicit in telling the player something which makes the actions make sense to them.


Of course, this isn't always wanted or beneficial, but it's definitely worth keeping in mind.

In Topic: How to get from A to Z (or my first dev blog)

25 September 2016 - 12:20 PM

BTW: If you do want to make a blog, there are developer journals here for exactly that purpose, the forums are just for discussions.

Expanding on this:

Developer journals can be found here: http://www.gamedev.net/blogs

You can create your own here (button to the upper right -- "Create a Journal"): http://www.gamedev.net/index.php?app=blog&module=manage

In Topic: Mentioning someone in a post

20 September 2016 - 08:02 AM

I didn't get any notification about that, at least.


But hello!

In Topic: Find angle to get from point A to B with x amount of bounces

20 September 2016 - 03:18 AM

You could shoot in 4 directions, not just 2.


Before morning caffeine, I believe you can do something like this...


In the selected direction, calculate the "effective" end point, as if this were just a huge box, with no bounces.

Basically, add get the distance to the border in your selected direction, multiply it by the amount of bounces, and add that to your endpoint.

Then just get the angle between the start point and that "effective" end point.


EDIT: You might have to add a different distance on odd/even bounces -- (absolute) distance to border in selected direction for odd bounces, (absolute) distance to border in opposite direction for even bounces.

So 5 bounces gives you an effective end point that is offset by (3*absolute_distance_to_border_in_selected_direction + 2*absolute_distance_to_border_in_opposite_direction).

EDIT 2: The previous edit wasn't quite correct -- which you can see in Nanoha's image below, but I think you're able to get a rough understanding of what you need to change in order for it to work :)

In Topic: How to detect diamond shape collision detection , if object is inside the dia...

20 September 2016 - 03:00 AM

Thanks! You mean pixel perfect collision detection?

Yes. Take a look at the following link for more information: